INDEX

Explanations

phrases related to instructional tutorials and coding languages

oai_token-act-pair · gpt-3.5-turbo

second-person pronouns and references to user capability or action

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

interstitial

-1.25

drug

-0.83

 condol

-0.76

 funer

-0.72

onement

-0.72

utral

-0.71

 Moody

-0.70

 Celebrity

-0.68

olit

-0.67

ELD

-0.67

POSITIVE LOGITS

 install

1.43

 configure

1.36

 automate

1.33

 migrate

1.26

 installing

1.25

 manually

1.18

 implement

1.18

 ported

1.18

 interoper

1.17

 installs

1.15

Activations Density 0.352%

phrases related to instructional tutorials and coding languages

second-person pronouns and references to user capability or action

No Comments

No Known Activations